lable at ScienceDirect

Quaternary International 364 (2015) 217e230

Contents lists avai

Quaternary International

journal homepage: www.elsevier .com/locate/quaint

Karst landscape evolution in the littoral area of the Bay of Biscay(north Iberian Peninsula)

Arantza Aranburu a, b, *, Martin Arriolabengoa a, b, Eneko Iriarte b, c, Santiago Giralt d,I~naki Yusta a, Virginia Martínez-Pillado a, b, Miren del Val a, b, Javi Moreno e,Montserrat Jim�enez-S�anchez f

a Department of Mineralogy and Petrology, Faculty of Science and Technology, University of Basque Country (UPV-EHU), Barrio Sarriena, s/n,48940 Leioa, Spainb ARANZADI Geo-Q Zentroa, Mendibile Auzoa, 48940 Leioa, Bizkaia, Spainc Department of Historical Science and Geography, University of Burgos (UBU). Edificio IþDþi, Plaza de Misael Ba~nuelos s/n, 09001 Burgos, Spaind Institute of Earth Sciences Jaume Almera (ICTJA-CSIC), Sol�e i Sabaris s/n, E-08028 Barcelona, Spaine Department of the Environment and Land Use Policy, Basque Government, Donostia-San Sebastian, 1, 01010 Alava, Spainf Department of Geology, University of Oviedo, Arias de Velasco s/n, 33005, Spain

a r t i c l e i n f o

Article history:Available online 23 October 2014

Keywords:Cone-type karstEndokarst sedimentAllostratigraphyLandscape evolutionPleistoceneCantabrian margin

* Corresponding author. Department of MineralogScience and Technology, University of Basque Countrys/n, 48940 Leioa, Spain.

E-mail address: arantza.aranburu@ehu.es (A. Aran

http://dx.doi.org/10.1016/j.quaint.2014.09.0251040-6182/© 2014 Elsevier Ltd and INQUA. All rights

a b s t r a c t

The western Pyrenean area contains extensive karst areas, however, their genesis and development arestill mostly unknown. In this work, we make a general description of the karst landscape in differentkarst areas: 1) Rasa type; 2) Cone-type karst; and 3) Alpine-type karst. The first two types are present inthe littoral area, where geomorphological evolution mostly depends on sea level and climatic changes.We focused our study in two karst areas of the littoral cone-type karst. We correlate different cavescreated from different stable water table levels. Three representative caves are studied in detail, studyingtheir stratigraphic record based on allostratigraphy. Finally we dated the different phases of speleothemformation in the three caves. Using all this information, we defined four stable paleowater table levels, at50, 150, 220 and 350 m asl, controlled by sea level changes and isostatic uplift events. The lowest level isthe youngest, with an age of c. 1 Ma. We discovered that the interior of the three caves display verysimilar endokarst allostratigraphic sequences, characterized firstly by an erosion phase, a fluvio-karstinput, flowstone speleothem formation and finally dripping speleothem formation. The chronologicaldata shows a correlation between these phases and Pleistocene climatic phases; the erosion phase isrelated to the falling sea level, fluviokarst detrital input is related to cold (glacial) stages and the for-mation of speleothems is related to the warmest (interglacial) moments and high sea level periods.

© 2014 Elsevier Ltd and INQUA. All rights reserved.

1. Introduction

Surface and subsurface karst morphologies are key componentsfor understanding the nature and genesis of cave and karst systems(De Waele et al., 2009). Each karst system has many differentcontrols, including the regional climate and the mechanical andchemical composition of the carbonate host-rock. Over time, cavesystems normally tend to evolve towards water-table caves, ascomputer modelling has confirmed (Gabrov�sek and Dreybrodt,2001; Kaufmann, 2009). Horizontal karst level development istherefore considered to be a product of a stationary paleo-water

y and Petrology, Faculty of(UPV-EHU), Barrio Sarriena,

buru).

reserved.

table (B€ogli, 1980; Ford and Williams, 2007; Audra and Palmer,2011) and can be identified by speleogenetic features that indi-cate the transition from phreatic to vadose conditions (Palmer,1991; H€auselmann, 2002) when the base level falls. As the water-table descends, due to the effect of a fall in the base level (sealevel or tectonic uplift), phreatic horizontal caves become vadoseand cyclical repetition of this process forms a multi-level karstsystem (Palmer, 1987; Granger et al., 2001; Anthony and Granger,2007; Strasser et al., 2009; De Waele et al., 2012).

Moreover, caves are natural sediment traps, and yield valuablepaleoenvironmental and geomorphological information (Richardsand Dorale, 2003; Sasowsky, 2007), especially in areas whereexternal landforms are predominantly erosive. Three different endo-karst processes eallochthonous sediment input, sediment erosionand chemical sediment precipitatione are related to distinct

A. Aranburu et al. / Quaternary International 364 (2015) 217e230218

paleoclimatic conditions that appear to have occurred repeatedly inthepast, formingcomplexendokarst sedimentary records(Auleret al.,2009). In karst sedimentary sequences, correlation of unconformitiesthroughout the karst cavities is a basic tool for delimiting and inter-preting different endokarst sedimentary sequences and/or units,especially when they have a similar composition (Hughes, 2010).

Karst landscape is widespread in the BasqueMountains but onlyfew modern studies offer general information about their charac-teristics and forming processes (Maiztegi et al., 1974; Ugarte, 1985,1989; Era~na and Ugarte, 1992). The karst features vary greatly fromthe littoral to the inner highlands, allowing three landscape unitsfrom N to S to be defined: 1) The uplifted marine terraces or coastalplains on top of carbonate rocks that outcrop discontinuously alongthe coastline show a very subtle exokarst modelling and a relativelyhigh endokarst development (Moreno et al., 2010; Jim�enez-

Fig. 1. Location and geological maps of the studied areas. The location of the three stud

S�anchez et al., 2011b); 2) The cone-doline karst landscape, whichdeveloped in the inner littoral to highland transitional area up to350 m asl, where the landscape is made up of isolated roundedconical hills (cone karst or Fengcong karst, Waltham, 2009) andextensive development of sinkholes and small poljes or karst val-leys; and 3) the Alpine-type karst, located in the inner highlandswith very steep reliefs, where the vertical waterflow is dominant.

The three karst units aremodelled on the same lithologye LowerCretaceous shallowwater limestones. The different geomorphologicfeatures observed in each karst landscape unit may therefore beindicative of the influence of factors other than bedrock lithology,such as i) different emersion chronology, ii) the superimposition oftectonic, eustatic or geomorphologic processes over time (polygenicmodelling), and iii) the differential paleoenvironmetal impact ofQuaternary paleoclimatic changes at different altitudes.

ied caves is indicated: Goikoetxe (Urdaibai), Praileaitz (Deba) and Urtiaga II (Deba).

A. Aranburu et al. / Quaternary International 364 (2015) 217e230 219

This research focuses on the littoral cone-doline karst typelandscape located in the Cantabrian margin of the Basque Moun-tains (N Iberian Peninsula) (Fig. 1). The aims of this work are 1) toset out initial data on the geomorphologic and geochronologicalcharacterization of this karst landform; 2) to understand thedifferent depositional and erosive processes that control endokarstdevelopment and evolution and 3) to correlate allostratigraphicunits between different karst systems and the paleoclimaticchanges during the Pliocene-Quaternary.

2. Regional setting

The study areas are located in the littoral zone of the BasqueCountry, in the northern Cantabrian margin of the Iberian Peninsula(Fig. 1). The landscape shows subparallel NWeSE mountain rangeswith altitude increasing towards the SE (Ugarte, 1994). The highestsummits rise to 1500 m at a maximum distance of 70 km from thecoast and separate the Cantabrian (Atlantic) and Mediterranean wa-tersheds. The valleys have steep slopes and a very intense fluvial ac-tivity isdevelopeddue totheAtlantic-typetemperateandwetclimate,(10 �C annual average and 1200e2400 mm/y) with no dry season.

The region studied lies within the Mesozoic Basque-CantabrianBasin, in the Basque Arc structural domain (Feuill�ee and Rat, 1971),on the northern flank of the Bilbao anticlinorium. The telogenetickarst is developed on Lower Cretaceous (Aptian-Albian) stratifiedto massive shallow-water micritic limestones (Fig. 1), with rudistsand corals formed in warm tropical seas (Agirrezabala, 1996).

The Alpine orogeny caused deformation and emersion of thesemarine limestones, first of the southern continental area, duringthe Eocene e Upper Oligocene, and during the Miocene of thenorthern area (García-Mond�ejar et al., 1985). The continued tec-tonic uplift, the calcareous nature of the emerged rocks and thepresence of the first calcareous intramontane lacustrine basinssuggest that the process of karstification in the Cantabrian marginmay have started in the Miocene.

Fig. 2. Photographs of the main cone-doline karsts features in the study areas: A) Deba areathe valley is formed by a small karst valley through which the River Deba flows. The Praileatihill (350 m): the valley bottom is formed mostly by the estuarine plain. It hosts the Goikoeblind karst valley form the valley bottom. It includes the Urtiaga II cave.

The valleys of the Cantabrian margin are short and steep-sided.The relief is mainly controlled by the intense incision of the riversthroughout the Quaternary (del Val, 2013), whereas glacial land-forms are only present locally in the Basque Mountains in areaswith an altitude of over 1000 m (e.g. the Aralar range, Rodríguez-Rodríguez et al., 2015). The littoral area is characterized by asteep cliffy coast interrupted by fluvial valleys and Holocene estu-aries (Leorri et al., 2012). The western section of the Cantabriancoast has well-developed marine terraces (Flor, 1983; Mary, 1983;Alvarez-Marr�on et al., 2008; Jim�enez-S�anchez et al., 2011a), mostof which have been dismantled by fluvial erosion to the east, suchas the Basque coast.

2.1. Study areas

The features and geomorphological evolution of the karst land-scape in the littoral zone are concentrated in two different areas:Urdaibai and Deba (Fig. 1). Both areas contain a well-developedcone-doline karst (Fig. 2), a cliffy coast with dismantled planationsurfaces, and the formation of an estuary at the mouth of the mainrivers (Fig. 3). The karst area of Urdaibai is known as Busturialdea,with 31.22 km2 of limestone surface, 235 caves and 20.21 km ofendokarst development catalogued to date (ADES, 2010; Doradoet al., 2013). In Deba, there are two karst areas, Arno (12.68 km2)and Izarraitz, with 36.05 km2 of limestone surface, 236 caves and31.07 m of endokarst development catalogued (Dorado et al., 2013).

The endokarst analysis focuses on three caves that summarizethe most complete stratigraphy observed in the different studiedcaves within the cone karst areas (Fig. 2): Goikoetxe (Urdaibai area),Praileaitz (Deba area), and Urtiaga II (Deba area) (Figs. 1 and 3).These cavities were chosen for (i) their well-developed cave sys-tem, (ii) the presence of significant phreatic and vadose geomor-phological features and sedimentary record, (iii) the existence ofspeleological and archeological bibliography and (iv) theiraccessibility.

: the summit of the conical karst pinnacles are located around 120 m and the bottom ofz cave is located in the pinnacle that is being quarried; B) Urdaibai area, the Pe~na Foruatxe cave; C) Deba area, cone type hills at 215 m altitude: various sinkholes and a small

Fig. 3. LiDAR based digital elevation model (DEM) of the study areas. The colored areas correspond to the horizontal areas related to the potential planation surfaces. A) Urdaibaistudy area; B) Deba study area.

A. Aranburu et al. / Quaternary International 364 (2015) 217e230220

2.2. Studied caves

2.2.1. Urdaibai: Goikoetxe CaveThe karst system is over 3400m long (ADES, 2010) and is located

on the west flank of the diapiric anticline of Urdaibai, on the left

bank of the Urdaibai estuary (Fig. 1). It is situated in a 350 m highconical hill (Pe~na Forua), formed by Albian limestones oriented at60�/255� (EVE, 1992). It is bounded on the west by a normal faultwith siliciclastic rocks, and on the east by the Oka river estuary(Urdaibai).

A. Aranburu et al. / Quaternary International 364 (2015) 217e230 221

The principal network of NeS galleries comprises three sub-horizontal levels of phreatic origin (Fig. 4A). Structural control hascaused abrupt changes in the direction of the galleries and anenlargement of the cavity where various discontinuities coincide.The active karst level (30 m asl) receives autochthonous water fromthe recharge zone, marked by extensive development of sinkholes,and allochthonous water from the siliciclastic massifs, which drainthrough small water courses running laterally to the karst. Thiswater emerges through the Apraiz spring (Fig. 3). The intermediatelevel (50 m asl) constitutes 80% of the karst system, and includesabundant speleothem formations and terrigenous deposits(Aranzabal and Maeztu, 2011). In the upper level (80 m asl), whichis practically filled with sediment, it has been possible to exploreonly a few meters.

2.2.2. Deba: Praileaitz cavePraileaitz cave is located on the NE side of the Praileaitz hill

(currently used as a quarry) (Fig. 2), near the Deba river estuary(Figs. 2 and 5A). The river meanders around the Albian limestonehills that outcrop throughout the coastal strip in an N120�E direc-tionwith a subhorizontal or 20e30� N dip (Agirrezabala, 1996). The

Fig. 4. Goikoetxe cave: A) Longitudinal cave profile, in which the three karst levels can be se(modified from Aranzabal and Maeztu, 2011); C) Idealized cross-section, allostratigraphic u

hills with summits of around c. 120e150 m are conical in shape,partially covered by a limestone pavement, and separated by karstvalleys (Fig. 2). The cave entrance is located at 55 m asl (Fig. 5A). Ithas a subhorizontal development, with features of phreatic disso-lution and scallopmarks. It runs 100m in an N160e170� E directionparallel to one of the main fracture systems in the massif (Fig. 1).The cave is intersected by 3 groups of discontinuities which, amongother factors, are the causes of the main drip points and the for-mation of speleothems (Fig. 5B) (Iriarte et al., 2010). The cavity ispartially filled by sediments that in the upper part containarcheological levels of Musterian-Azilian age; Solutrean cavepaintings are also present in the cave walls (Pe~nalver and Mujika,2005). Other cave passages are known to exist at higher andlower elevations, but are blocked with sediment.

2.2.3. Deba: Urtiaga II caveThis karst system is located in a 215 m high cone-shaped hill of

Aptian-Albian limestones, located 1.6 km from the coast (Figs. 3 and6A). It is isolated from the rest of the limestonemassif by flysch siltysandstones and marls from the Upper Albian lithostratigraphic unit(Fig. 2) (EVE, 1992). To the south of the hill there is a small blind

en; B) Plan of the cave and detailed topography of the study area, called “the red room”

nits and location of dated samples in Goikoetxe cave.

Fig. 5. Praileaitz cave: A) Longitudinal profile of the cave; B) Detailed topography of the cave (modified from Iriarte et al., 2010); C) Idealized cross-section, allostratigraphic unitsand location of dated samples in the cave.

A. Aranburu et al. / Quaternary International 364 (2015) 217e230222

valley, formed in the terrigenous rocks, which drains to the karst(sump). This system has two types of water infiltration: autoch-thonous and dispersed infiltration, resulting from rainfall, andconcentrated infiltration of river water, through the sump (Gal�anet al., 2004).

Urtiaga-II is a 502m long branching system. The cave entrance isat 149 m asl. Three separate galleries (Main, Upper and River) canbe distinguished with NeS andWNW-ESE orientation (Gal�an et al.,2004) (Fig. 6B). TheMain Gallery is located at 145m asl, 30m abovethe active water-course. Stratigraphic evidence indicates that thecavity was nearly clogged by fluviokarst sediments that were sub-sequently eroded away. The cave contains a prehistoric paleonto-logical site (Altuna, 1995).

3. Methods

The surficial geomorphological features of the study areas wereanalyzed from digital elevation models (DEM) using LIDAR data(1 m resolution and 0.3 m vertical accuracy), taken into account thefield and speleogenesis data (De Waele and Parise, 2013), with theaim of identifying geomorphic features (morphometric indicators)which could be used as markers of the effects of interaction

between sea-level, groundwater and karst formations (Canoraet al., 2012). Dismantled potential planation surfaces were identi-fied through DEM processing, in a modification of the methodologyproposed in del Val (2013) for the identification of river terraces.The potential areas of former planation surfaces here consideredare areas with a slope of less than 13� and that do not coincide withvalley floor areas and which have not suffered severe humanalteration (anthropic origin).

Analysis of the endokarst development is based on the cata-logue of the Union of Basque Speleologists (www.euskalespeleo.com) and Asociaci�on Deportiva Espeleol�ogica Saguzaharrak(ADES, 2010). In order to locate ancient stable water table levelswithin the coastal karst areas, only caves with a significant hori-zontal development were considered.

These caves were selected using the following procedure: i)remove/disregard caves with less than 50 m of horizontal devel-opment; ii) remove caves that are now under sea level; iii) chooseonly caves from Busturialdea, Arno and Izarraitz karst areas; iv)remove caves with more than 20 m of vertical development; v)aggregate caves whose topography is known or public and aresuitable for this purpose. Finally, the selected caves were plotted ona graph as a function of their distance from the shoreline and

Fig. 6. Urtiaga II cave: A) Longitudinal profile of the Urtiaga II karst system, with the studied upper karst level shown in red; B) Detailed topography of the cave (modified fromGal�an et al., 2004); C) Idealized cross-section, allostratigraphic units and dated-samples location on the cave. (For interpretation of the references to colour in this figure legend, thereader is referred to the web version of this article.)

A. Aranburu et al. / Quaternary International 364 (2015) 217e230 223

topographic elevation (Fig. 7C). The aimwas to assess whether theycorrespond to different levels of endokarst development.

Based on accessibility and the completeness of the stratigraphicrecord, three caves corresponding to two different stationary paleo-water table levels were chosen and studied. In order to reconstructthe succession of geological events, cross-sections and stratigraphiccolumns were made and characterized in different areas of thecaves. The sedimentary infill was ordered into lithostratigraphicand allostratigraphic units (Hughes, 2010).

Fourteen absolute dates were obtained from speleothem dating(flowstones and stalagmites) in the three caves. The samples weredated using the Uranium-series disintegration (234U/230Th) methodproposed by Ivanovich and Harmon (1992), using BR-024-450-100ORTEC OCTETE PLUS alpha spectrometers at the Jaume AlmeraInstitute of Earth Sciences (ICTJA-CSIC). Chemical separation of the

radioisotopes and purification were performed following the pro-cedure described by Bishoff et al. (1988). Isotope electrodepositionwas carried out using the method described by Talvitie (1972) andmodified by Hallstadius (1984). Age calculations were based on thecomputer program by Rosenbauer (1991). Owing to the largeamounts of terrigenous clay content in three samples, ages werederived after applying the isochron technique (Bischoff andFitzpatrick, 1990).

Finally, we have integrated other dates of speleothems availablein the literature to compare and complete the speleothem forma-tion stages observed in our study. The data are obtained from caveslocated on the Atlantic coast (littoral karst) (Jim�enez-S�anchez et al.,2011a), along the Cantabrian margin of the Pyrenees (Vanara,2000), and in the Central Pyrenean and the Iberian Massif(Moreno et al., 2012).

Fig. 7. Profiles of the horizontal areas corresponding to the potential planation surfaces shown in Fig. 3, for A) Urdaibai area and B) Deba area. C) Projection of main subhorizontalcaves in multilevel karst systems from the studied littoral areas. The vertical bar shows the altitude variation of the considered cave (max. ±20 m), while the horizontal bar indicatesthe development of each cave.

A. Aranburu et al. / Quaternary International 364 (2015) 217e230224

4. Results

4.1. Cone-doline karst: multiple phreatic stabilization karst leves &planation surfaces

BasedonprocessingofDEMdata, 5 levelshavebeen identifiedwiththe greatest development of horizontal surfaces (potential ancientplanation surfaces) in Urdaibai (Figs. 3A and 7A) and 3 in Deba karstareas (Figs. 3B and 7B). In Urdaibai these are located at the followingaltitudes: 25e55 m, 65e80 m, 100e145 m, 220e235 m, and270e330 m (including the Goikoetxe cone-hill). In Deba they are ataltitudes of: 100e180 m (including the Praileatiz cone-hill),200e240 m (including the Urtiaga II cave cone-hill), and290e340 m. These three levels at Deba area coincide with the threeupper levels of Urdaibai, whereas the lower two levels from Urdaibaiare absent in Deba area.

The karst landscape generated by dissolution from the threehighest planation surfaces displays conical pinnacles (Fig. 7A, B and2) with well-developed dissolution features such as karren, sepa-rated by dolines and small karst valleys, and not snow-forms andcolluvial deposits are observed. The valley floor nearby the studiedcaves is at 50e5 m asl (Praileaitz), 145e90 m asl (cone-hill ofUrtiaga II), and 80e30 m asl (Pe~na Forua). This karst landscapeevolution could correspond to Stage 2 or 3/4 of Waltham (2009).

The horizontal surfaces might represent former subhorizontalerosion surfaces (planation surfaces), fromwhich karstification andfluvial incision began each time that they were rejuvenatedthrough a rapid tectonic uplift (Ford andWilliams, 2007; Waltham,2009). However the original planar erosive surfaces of the studiedareas are strongly eroded, leaving small horizontal remnants in thepresent landscape (Figs. 3 and 7).

At the same time, the height of subhorizontal cavities of thelittoral karst massifs of Urdaibai, Izarraitz and Arno reveal six roughlevels of endokarst development around 20 m, 50e70 m, 110 m,160 m, 205e265 m and 350e380 m asl (Fig. 7C). Both Urdaibai andDeba area caves are found in all of the levels, except in the highestone, where only Urdaibai caves are found.

The altitudes of the planation surfaces roughly coincidewith thealtitude of the main subhorizontal cavity system arrangement inobserved karst levels. The two lower levels of the multi-level karstsystem can be correlated in height with the two lower horizontallevels of Urdaibai surfaces, whereas in Deba area these levels areabsent. Karst levels around 100e160m and 200e260m are presentin the three graphs (Fig. 7). However, the highest level of multi-level karst is around 360 m, while in Urdaibai planation surface ispresent between 270 and 330 m and in Deba it is at 290e340 m.This may be due to the fact that the highest planation surfaces(above 360 m) are more eroded.

4.2. Allostratigraphy of cave sediments

4.2.1. Urdaibai: Goikoetxe Cave (50 m asl)The erosive discontinuities are the best correlation features

throughout the intermediate karst level (Fig. 4C). The sedimento-logical record of the intermediate karst level (Goikoetxe Cave) in-cludes two allostratigraphic units (Aranburu et al., 2011).

4.2.1.1. Allostratigraphic unit-1. Throughout the entire cave, thereare often allochthonous siliciclastic detrital sediments, up to 5 mthick, resting on the limestone conduit. These are mainly siliceoussands and gravels (1e10 cm) derived from siltstone, claystone,sandstone and limonite nodules (Figs. 4B and C). The sedimento-logical features suggest a polyphasic transport of sediments, carriedby river systems, from the siliciclastic bedrock that outcrops to theE of the limestone hill (lateral recharging). This fluvio-karst infillclogged up nearly the entire phreatic cavity.

The detrital sequence culminates with the development of avery compact subhorizontal flowstone, 40e67 cm thick.

4.2.1.2. Allostratigraphic unit-2. The subsequent reactivation of thecave caused the partial erosion of the previous sedimentary infill,leaving a relict stratigraphical sequence in the cave. The erosivenature of this process is noteworthy, causing the removal of 3e4 m

A. Aranburu et al. / Quaternary International 364 (2015) 217e230 225

in thickness of detrital and chemical sediments, with the devel-opment of a paleorelief.

A second generation of speleothems, mainly formed of stalac-tites and stalagmites, developed densely along the erosion surfaceand is still active. In some areas of the gallery, speleothems areaffected by gravitational collapses. In collapsed areas, speleothemgrowth has continued to generate a new ensemble of stalagmiteson the collapsed blocks (Fig. 4C).

On the west side of the gallery, in which first generation flow-stone did not grow, there is a flowstone up to 2 m, coeval with thesecond speleothem generation or below the third (Fig. 4C), asdeduced by its stratigraphic position.

4.2.2. Deba: Praileaitz Cave (55 m asl)Two thirds of the volume of the phreatic cavity is filled with

endokarst sediment. It has a minimum stratigraphic thickness of4 m, and 7 sedimentary phases have been differentiated, with apredominance of chemical precipitation over detrital sedimenta-tion in a ratio of 5:2. The sedimentary succession can be groupedinto two allostratigraphic units (Fig. 5C), from bottom to top:

4.2.2.1. Allostratigraphic unit-1. The oldest stratigraphic leveloutcropping in the phreatic conduit is made up of stalagmites andflowstones located above the present floor of the cave. In someareas of the cave, the flowstone has an intercalation of terrigenousgravels: calcite-cemented quartzitic and limonitic pebbles less than3 cm in diameter, with very little sandy matrix.

4.2.2.2. Allostratigraphic unit-2. The beginning of this unit ismarked by major erosion, which left remnants of previous flow-stone and conglomerates in the rock walls 1 m above the presentcave floor. The paleorelief was filled with allochthonous siliciclasticdetrital sedimentation, with an average thickness of 2 m. The upperpart of these sediments includes archeological levels ranging fromthe Solutrean to the Azilian (Pe~nalver and Mujika, 2005). An8e20 cm thick flowstone covers the detrital sediments. Above theflowstone, Holocene stalagmites and stalactites have developed.

4.2.3. Deba: Urtiaga II Cave (150 m asl)The endokarst sedimentary record in Urtiaga II Cave can be

summarized in three allostratigraphic units. The endokarst sedi-ments vary in thickness between 3 and 5 m, and are formed by thealternation of siliciclastics and speleothems (Fig. 6C).

Table 1Speleothem samples dated using the 234U/230Th method. The ages obtained cover a lonsamples (Pra_3 and Gk_3) were in secular equilibrium owing to their age beyond the lim

Sample Ref. Lab U238 (ppm) Th-232 (ppm) U-234/U-238 Th-230/Th-2

Gk_1 2012 0.18 **** 0.81 ± 0.02 ****Gk_2 2112 0.31 **** 0.81 ± 0.02 ****Gk_3 bulk 2212 0.09 0.15 1.11 ± 0.04 2.096 ± 0.10GK_3 lix 3712 0.09 0.28 1.23 ± 0.06 1.266 ± 0.04GK_3 res 3912 1.26 5.07 0.96 ± 0.13 0.661 ± 0.08Gk_4 2312 0.26 0.65 0.97 ± 0.03 1.080 ± 0.02Gk_5 lix 2412 0.16 0.19 0.98 ± 0.03 2.052 ± 0.10Gk_5 lix 3512 0.17 0.30 1.16 ± 0.04 1.432 ± 0.06Gk_5 res 3612 1.75 4.20 0.78 ± 0.07 0.836 ± 0.09Gk_6 2512 0.31 0.25 0.87 ± 0.02 2.172 ± 0.10Pra_1 113 0.03 0.08 1.21 ± 0.09 0.846 ± 0.06Pra_2 213 0.03 0.06 1.32 ± 0.11 1.258 ± 0.10Pra_3 313 0.04 0.11 1.35 ± 0.1 1.614 ± 0.13Urt_1 513 0.33 0.04 1.12 ± 0.02 26.89 ± 2.14Urt_2 613 0.92 0.1 1.21 ± 0.01 14.596 ± 0.6Urt_ 3_bulk 1613 0.35 0.51 0.98 ± 0.02 2.577 ± 0.07Urt_ 3_lix 2913 0.36 0.29 1.36 ± 0.03 4.282 ± 0.18Urt_ 3_res 3013 3.12 6.32 0.75 ± 0.04 0.83 ± 0.07

4.2.3.1. Allostratigraphic unit-1. The oldest stratigraphic levelcomprises different types of speleothems (flowstones, stalactites,draperies, etc.) strongly eroded by an erosive surface.

4.2.3.2. Allostratigraphic unit-2. The erosive surface is covered byallochthonous siliciclastic sediments, 1e3 m in thickness. There is apredominance of <5 cm well rounded pebbles of sandstone andsiltstone. These are arranged in 0.4 m thick beds, with massiveinternal structure or parallel lamination. They are interpreted asfluviokarst facies derived from the transport of allochthonoussediment by the river from the external blind valley.

In the upper part of the fluvial sequence, there is a 3e5 cm thickflowstone. The formation of this flowstone, overlying pebbles, in-dicates a decrease of thewater flow, with a predominance of diffuseinfiltration of karst waters. The sequence finishes with the forma-tion of a paleontological site, with abundant remains of cold-climate fauna (Altuna, 1995), and the formation of another deci-metric flowstone on top covering the whole surface.

4.2.3.3. Allostratigraphic unit-3. The erosion and removal of almost2 m of the previous endokarst infilling marks the beginning of thethird allostratigraphic unit. The steep paleorelief is fossilized by finesediments (silt and clay), probably carried by the water from theentrance of the cave and/or by infiltration. These fine sedimentscontain the hibernation nests of bears (the species is not known)and Holocene stalagmites.

4.3. Chronology

The chronology of the cave sediments is based on speleothemdating and the archeological and paleontological data available.Results of the U-series ages are shown in Table 1. In general, 238Ucontent in the dated samples is low (mean value 0.26 ppm) andmost of the samples are not pure carbonates but contain variableamounts of clay owing to their respective 230Th/232Th ratio, rangingfrom 0 to 26.89. Of the 12 dated samples, only two (Gk_1 and Gk_2)were composed of pure calcite with no terrigenous contaminationtraces, and twomore (Urt_1 and Urt_2) contained almost negligiblepresence of detrital particles (230Th/232Th ratio > 14). However, theage of three of the contaminated samples (larger 232Th amount)were successfully corrected by isochron plots. The other five sam-ples contained detrital material and the nominal ages obtainedshould only be considered as approximate.

g period of time, from 7000 ± 520 years (Gk_2) to 278,000 years (Gk_5). Only twoit of the method.

32 Th-230/U-234 Nominal date (years BP) Isochron date (years BP)

0.12 ± 0.01 14,548 þ 1010/�10000.06 ± 0.00 7000 þ 520/�520

3 1.04 ± 0.05 >350,0008 1.04 ± 0.05 >350,0001 0.89 ± 0.12 >350,000 >350,0007 0.91 ± 0.03 278,250 þ 64,985/�40,3955 0.78 ± 0.04 165,328 þ 19,475/�16,4958 0.71 ± 0.03 128,762 þ 11,318/�10,3116 0.83 ± 0.09 >350,000 139,546 þ 31,839/�24,8196 0.65 ± 0.03 118,055 þ 9890/�90203 0.75 ± 0.06 141,074 þ 23,621/�19,6692 0.66 ± 0.05 112,260 þ 16,384/�14,4205 1.09 ± 0.08 >350,0006 0.95 ± 0.03 272,831 þ 39,390/�29,39141 0.42 ± 0.01 57,785 þ 1898/�1868

1.24 ± 0.04 >350,0006 0.81 ± 0.03 159,995 þ 12,865/�11,645

0.83 ± 0.07 >350,000 126,464 þ 23,634/� 19,998

A. Aranburu et al. / Quaternary International 364 (2015) 217e230226

The dates obtained from the three studied caves and othersobtained from previous works (see Fig. 8) are clustered in differentperiods of time or generations. Four separate generations of spe-leothem formation have been identified:

� Speleothem Generation-I, the oldest one, is flowstone type anddates from an interval between >350 and 273 ka, around theMIS 9 (Fig. 8). It is present in the three studied cavities, and it isincluded within the allostratigraphic unit-1 in all three cases. Itis also present in Asnarre and Pindal littoral caves (Jim�enez-S�anchez et al., 2011a and R�eseau du N�eeb�el�een cave (Arbailles,Vanara, 2000).

� Speleothem Generation-II is also present in the studied threecaves as flowstone type, except at Goikoetxe cave wheredripstones formed. The growth period for this phase is be-tween 141 and 112 ka (MIS 5e6). The speleothems fromGeneration-II are included within the allostratigraphic unit-3 ofUrtiaga II cave. In Pindal cave speleothem from MIS 5e-MIS 6has also been recorded (Jim�enez-S�anchez et al., 2011a, 2011b),while in Arbailles (Vanara, 2000), central Pyrenees and centralnortheast of the Iberian Peninsula MIS 5 (Moreno et al., 2013)(Fig. 8).

� Generation-III is flowstone type and formed c. 58 ka (MIS 3). It isfound only in the Urtiaga II cave, carpeting the detrital sequenceat the end of the allostratigraphic Unit 3. This generation ap-pears also in other caves in the central and part of the IberianPeninsula (Moreno et al., 2013).

� Generation-IV speleothems are Holocene (MIS 1). In Prai-leaitz cave, this generation has been dated by its stratigraphicposition above the Azilian levels (Early Holocene) (Pe~nalverand Mujika, 2005), consisting a flowstone that evolves tostalactites/stalagmites. This speleothem phase is also presentin Pindal cave (Jim�enez-S�anchez et al., 2011b), Grotted0Azal�eguy in north Pyrenees (Vanara, 2000) and centralnortheast of the Iberian Peninsula (Moreno et al., 2013).

Fig. 8. Speleothem chronology and growing phases during the interglacial stages. b) Proposeduring the fall and lowstand, siliciclastic sedimentation during the rise and speleothem formet al. (2006) and Marine Isotopic Stages (MIS) from Shackleton et al. (2003) are shown for

5. Discussion

5.1. Planation surfaces

DEM data suggests that there are 5 planation surfaces where thekarstification and fluvial incision began. In the case of the cone-doline karst relief of the study area, they are inferred to be theresult of dissolution affecting the planation surfaces located at100e150 m, 200e240 m and 300e350 m altitude (Fig. 7). Giventhat they have no associated sedimentary deposits, the nature ofthese erosion surfaces is not easy to determine. In the NE Canta-brian coast, Mary (1983), Flor (1983), �Alvarez-Marr�on et al. (2008),Jim�enez-S�anchez et al. (2006, 2011a) and Pedoja et al. (2014),interpreted that the strandplains formed in the littoral zone(8e10 km from the seashore) are of marine origin. In the coast ofAsturias (Jim�enez-S�anchez et al., 2006; �Alvarez Marr�on et al., 2008)the elevation of the emerged marine terraces (rasas) largelycovered byweatheredmarine and continental sediments, coincideswith the planation surfaces of our study area.

The proximity of the coast to the study area means that thewater table depends primarily on sea level. The fact that theplanation surfaces in the cone-doline systems coincide relativelywell with the stable paleo-water table levels of the karst systems(Fig. 7) suggests that their formation might have been related torelative sea level changes, reinforcing the hypothesis that theplanation surfaces might be previous marine abrasion platforms.

Eustasy caused by variations in water volume has clearlycontributed to this scenario, at least in the last 5 Ma (Lisiecki andRaymo, 2005). However, models suggest that the sea level duringthe interglacial maxima did not exceed the present height by morethan a few meters for the last 2.9 Ma (Bintanja and Van de Wal,2008), and possibly around 25 m during the Mid-Pliocene WarmPeriod (2.9 Ma to 3.3 Ma) and earlier (Raymo et al., 2011). Duringthe Pliocene and Lower Pleistocene, sea-level oscillations wereseemingly of smaller amplitudes and faster wavelengths (40 ka

d relationship between allostratigraphic units and the sea-level variation curve: erosionation during high sea level. The Relative Sea Level variation curve, taken from Rabineaureference. Most are situated between MIS 1 and MIS 5-6.

A. Aranburu et al. / Quaternary International 364 (2015) 217e230 227

cycle) and the abrasion platforms were more compact and mergeeasily into rasas (Pedoja et al., 2014). The uplift of the marineabrasion platforms and rejuvenation of the relief has therefore beencaused, at least in part, by tectonics.

The changes in the stable water tables in the Cantabrian coastoccurred due to isostatic processes, characteristic of a passivemargin (�Alvarez-Marr�on et al., 2008; Pedoja et al., 2011, 2014). Theupper level is therefore the oldest and the lower the most recent.The coincidence of elevations between the multi-level systems ofthe littoral karst, in the first instance, and also with the elevation ofthe paleo-abrasion platforms on the Cantabrian coast suggests atectonic uplift on a regional scale, without ruling out local move-ments throughout the Cantabrian coast (�Alvarez-Marr�on et al.,2008).

5.2. Cone-doline karst formation and development

Karst formation age can be determined from the maximum ageof the endokarst infills and the morphostratigraphy. In our case,there is no development of cone karst above the youngest þ50 mupliftedmarine terrace (rasa). This fact gives aminimum age for thelittoral karst. Taking into account that the þ50 rasa probablyemerged around 1Ma, dated by cosmogenic isotopes of the over-laying sediments (�Alvarez-Marr�on et al., 2008), and that the U/Thage for the oldest speleothems from caves developed within theuplifted strandlines 339 ka (Jim�enez-S�anchez et al., 2011a), this is aminimum age for the start of the formation of the cone-doline karstfrom planation-surfaces located at 150 m, 220 m and 350 m.

The rate of tectonic uplift inferred for the passive margins ofWestern Europe is between 0.06 mm/yr during MIS5e and0.08 mm/yr during MIS11 (Pedoja et al., 2014). Examples from theCantabrian margin indicate uplift rates of between 0.06 and0.15 mm/yr based on UeTh dating and cave incision (Jim�enez-S�anchez et al., 2006, 2011a) or 0.07 and 0.15 mm/yr dating theexposure and elevation of marine terraces with cosmogenic nu-clides (�Alvarez-Marr�on et al., 2008). Therefore, and assuming aconservative average uplift rate of 0.08mm/yr for the eastern Bay ofBiscay, the start of cone-type karst on the þ350 m rasa could bedated to approx. 4.3 Ma (between 4.3 and 2.3 Ma, Pliocene) (Fig. 9).In the northern Pyrenees, similar karst forms have been attributedto the Plio-Quaternary uplift (Vanara, 2000).

5.3. Endokarst record: stratigraphy and evolution

Each of the caves studied is located within a different cone-typekarst, formed from planation surfaces at different elevations.However, Goikoetxe and Praileaitz belong to the same stable paleo-water table (50 m asl), coinciding with the rasa at 50 m, whereasUrtiaga II stands at 150 m (asl).

Despite the geographical differences, the sedimentary record ofthe three caves shows several allostratigraphic units with a repet-itive and very similar depositional model: a) Erosive phase,marking the origin of the allostratigraphic unit; b) Allochthonoussiliciclastic sedimentation; and c) Formation of flowstone thatevolves to dripstones. This depositional model reflects a decrease inwater energy in the karst system, from the erosive phase(maximum energy) to the formation of speleothems from drippingwater.

The detailed sedimentary record varies depending on theparticular conditions of the environment of each cave. However, theconstituent phases of the allostratigraphic units are always thesame and, moreover, coeval. Three of the four generations of spe-leothem formation are present in all the caves studied and, ac-cording to the ages obtained, most of them were formed mainlyduring temperate periods. We are aware of the limited amount of

data available at this time, but comparisonwith data available fromthe littoral karst of the Atlantic coast (Jim�enez-S�anchez et al.,2011a), and even from the interior of the Iberian Peninsula(Moreno et al., 2013) support the validity of the generations ofspeleothems described in this paper (Table 1, Fig. 8).

Generation-I formed before 273 ka (c. MIS 9); Generation-II at141e112 ka (MIS 6-5); Generation-III (only in Urtiaga II), between60 and 56 ka (MIS 3) and generation-IV, during the Holocene (MIS1) (Fig. 8). These phases of speleothem formation coincide withthose described for the El Pindal Cave in the Cantabrian margin(Jim�enez-S�anchez et al., 2006, 2011b), the Arbailles karst massif(Vanara, 2000), both examples located around the Cantabrian sea,and alsowith the phases defined in the northeast central area of theIberian Peninsula (Moreno et al., 2013).

Almost all analyzed speleothem samples have a short growthtime, sometimes discontinuous, and are generally associated withan interglacial period. The oldest time interval in the formation ofspeleothems occurred between >350 and 270ka (>MIS9). Thispattern of speleothem growth is not observed during MIS 7 in thestudied caves, although this may be due to the limited number ofsamples. The time interval most favorable for the formation ofspeleothems occurred during the end of MIS 6-MIS 5, probablycoinciding with the MIS5e. This extensive phase is represented inall the studied caves and the Pindal cave in Asturias (Jim�enez-S�anchez et al., 2011b), but it appears relatively later in the north-east area of the Central Iberian Peninsula (Moreno et al., 2013),where optimal conditions were reached later due to the largeextent of glaciers during MIS 6 in the Pyrenees (Pe~na et al., 2004).Since MIS 5 there were not optimal conditions for the growth ofspeleothems in the caves of the Cantabrian coast (MIS 5a-MIS2).During MIS 3 there was only a brief growth of speleothems coin-ciding with the warmest interval of this isotopic stage which waspunctuated by rapid climate changes such as Heinrich events andDansgaardeOeschger cycles. The MIS 3-2 was probably too coldand arid, as well as being highly unstable, as evidenced by otherspeleothem records (Moreno et al., 2010), lacustrine sequences(Moreno et al., 2012) and glacial deposits (Lewis et al., 2009; Garcia-Ruiz et al., 2010) from northeast Spain. In the study area, wetterconditions established during the Younger Dryas, as deduced fromfluvial and alluvial records, both in the Pyrenees (Lewis et al., 2009)and in the Iberian Range (Fuller et al., 1998), and thus in the EarlyHolocene a new speleothemic phase (IV) started.

The allostratigraphic sequence of the caves shows a repetitiveinternal organization. The final stages in speleothem formationshow abrupt erosion features, with input of detrital sedimentscoinciding with the onset of the following allostratigraphic unitwhichappear tohavebeendepositedduring cold isotopic stages andhence a low sea level (Fig. 8B). Clastic sediment input to caves iscontrolled by the availability of sediments in the source area, sea-sonal availability of water and an increase in the potential energythrough falls in sea level and the presence of less vegetated soils, alltypical of cold events, related to cold episodes within the warmerperiod or with cold periods. The absence of dates in siliciclasticsediments does not enable us to know more about these shorterperiods. From the Middle Pleistocene to the present day, the peri-odicity and amplitude of the sea levels is around 100 ka and 120 m(Pedoja et al., 2014), causing great changes in the fluvial potentialenergy, at least in coastal zones. All these processes could be re-flected in the cyclic stacking of allostratigraphic units of the studiedendokarst sedimentary recordsof the littoral cone-doline type karst.

6. Conclusions

In the studied area of the Cantabrian margin, cone-doline karstis a landscape unit limited to the littoral zone between 5 and 10 km

Fig. 9. Simplified evolution and main processes involved in the formation of cone-type karst over time and current landscape forms in the eastern Cantabrian margin.

A. Aranburu et al. / Quaternary International 364 (2015) 217e230228

from the coastline and 100e350 m in altitude. This karst landscapeis developed starting from three different planation surfaces,interpreted as former marine terraces (rasas), tectonically upliftedto altitudes of 350, 220 and 150 m asl.

The start of the cone karst formation, inferred from the mor-phostratigraphic correlationwith marine terraces in the Cantabriancoast, the age of the older cave infills and a conservative rate oftectonic uplift (0.08 mm/y), could be traced back to during thePliocene, with a maximum development between the Pliocene andlower Pleistocene (4.3e2.3 to 1 Ma). The cavities generated in thecone-doline karst show multiple cave levels formed in stable watertable conditions. These cave levels coincide in altitude with theplanation surfaces (marine terraces, in origin). Changes in therelative sea level, due to tectonic uplift, and thus, of the

hydrological base level, seem to be the driving forces in the evo-lution of the karst landscape.

The lower and therefore most recent vadose karst level (þ50 m)displays cave sediments dated from >350 ka to the present. Thesesediments are grouped into two to three allostratigraphic units,which have a repetitive internal organisation: strong erosion,entrance of allochthonous siliciclastic sediment and speleothemprecipitation.

The genesis of these cave sediment sequences is closely relatedto the eustatic fluctuations caused by climate variations. Thus thefall in the sea level (cold periods) causes an increase in the potentialenergy of the rivers and therefore, the erosion and/or incision of thebase levels (lowstand). Availability of the sediment might havebeen favored by scant vegetation. Sedimentation of the

A. Aranburu et al. / Quaternary International 364 (2015) 217e230 229

allochthonous input of the fluviokarst system occurs as the sealevel rises (transgression). In warmer periods with less potentialenergy (high sea level), and reduced or better-regulated surfaceflow due to the presence of higher vegetation cover and activity, theinfiltration water becomes more acidic at first and the dissolvedcarbonate precipitates in the caves as speleothems. The mostfavorable time interval for the formation of speleothems occurredduring the end of MIS 6-MIS 5, coinciding with the period MIS5eEemian, and the Holocene (MIS 1). They are also present duringMIS3 and the end of MIS8-MIS9.

Acknowledgements

This research was funded by the University of the BasqueCountry's projects (UPV/EHU11/21 and UFI11/09). M. Arriolabengoaand M. del Val are funded by predoctoral grants (BFI-2012-289 andBFI-2010-379) from the Basque Government. The authors aregrateful to ADES and Aranzadi Speleology Groups for their coop-eration in the fieldwork. Warm thanks to Tim Nicholson for hiscontribution to correcting the English text. We are indebted with JoDe Waele and an anonymous reviewer for their helpful remarks.Their comments have considerably improved the originalmanuscript.

References

ADES, 2010. Cat�alogo de Cuevas y Simas de Urdaibai. Gobierno Vasco, Vitoria-Gasteiz.

Agirrezabala, L.M., 1996. El Aptiense-Albiense del Anticlinorio Nor-Vizcaino entreGernika y Azpeitia. Unpublished PhD thesis. University of the Basque CountryEHU/UPV.

Altuna, J., 1995. Urtiagako Leizea. In: Altuna, J., Armendariz, A., Etxeberria, F.,Mairezkurrena, K., Pe~nalver, X., Zumalabe, F.J. (Eds.), Carta arqueol�ogica deGipuzkoa. II, Cuevas. Munibe (Antropologia-Arkeologia), Suppl. o n�10.

�Alvarez-Marr�on, J., Hetzel, R., Niedermann, S., Men�endez, R., Marquínez, J., 2008.Origin, structure and exposure history of a wave-cut platform more than 1 M.a.in age at the coast of northern Spain: a multiple cosmogenic nuclide approach.Geomorphology 93, 316e334.

Anthony, D.M., Granger, D.E., 2007. A new chronology for the age of Appalachianerosional surfaces determined by cosmogenic nuclides in cave sediments. EarthSurface Processes and Landforms 32, 874e887.

Aranburu, A., Iriarte, E., Yusta, I., Giralt, S., Martinez-Pillado, V., Renteria, I.,Suarez, A., Arriolabengoa, M., 2011. Estudio paleoambiental a partir de precip-itados químicos: espeleotemas de la Sala Roja (Cueva Goikoetxe, Busturia,Bizkaia). In: Uni�on de Espele�ologos Vascos (Ed.), La cueva de Goikoetxe y elkarst de Pe~na Forua, pp. 101e118.

Aranzabal, G., Maeztu, J.J., 2011. El sistema Malloku y el karst de Pe~na Forua (Bus-turia, Bizkaia): la aportaci�on espeleol�ogica al proyecto Goikoetxe de custodiadel territorio. In: Uni�on de Espele�ologos Vascos (Ed.), La cueva de Goikoetxe y elkarst de Pe~na Forua, pp. 12e42.

Audra, P., Palmer, A., 2011. The pattern of caves: controls of epigenic speleogenesis.G�eomorphologie: relief, processus, environnement 4, 359e378.

Auler, A.S., Smart, L., Wang, X., Pil�o, B., Edwards, L., Cheng, H., 2009. Cyclic sedi-mentation in Brazilian cave: mechanisms and paleoenvironmental significance.Geomorphology 106, 142e153.

Bintanja, R., Van de Wal, R.S.W., 2008. North American ice sheet dynamics and theonset of 100,000-year glacial cycles. Nature 454, 869e872.

Bischoff, J.L., Fitzpatrick, J.A., 1990. Uranium series dating of impure carbonates: anisochron technic using total sample dissolution. Geochimica et CosmochimicaActa 55, 534e554.

Bishoff, J., Juli�a, R., Mora, R., 1988. Uranium-series dating of the Mousterian occu-pation at Abric Romaní, Spain. Nature 332, 68e70.

B€ogli, A., 1980. Karst Hydrology and Physical Speleology. Springer-Verlag, Berlin.Canora, F., Fidelibus, D., Spilotro, G., 2012. Coastal and inland karst morphologies

driven by sea level stands: a GIS based method for their evaluation. EarthSurface Processes and Landforms 37, 1376e1386.

De Waele, J., Plan, L., Audra, P., 2009. Recent developments in surface and subsur-face karst geomorphology: an introduction. Geomorphology 106, 1e8.

De Waele, J., Ferrarese, F., Granger, D., Sauro, F., 2012. Landscape evolution in theTacchi area (central-east Sardinia, Italy) based on karst and fluvial morphologyand age of cave sediments. Geografia Fisica e Dinamica Quaternaria 35,119e127.

De Waele, J., Parise, M., 2013. Discussion on the article “Coastal and inland karstmorphologies driven by sea level stands: a GIS based method for their evalu-ation” by Canora F, Fidelibus D and Spilotro G. Earth Surface Processes andLandforms 38, 902e907.

del Val, M., 2013. Utilizaci�on de los Modelos Digitales del Terreno a partir de datosLiDAR como instrumento de an�alisis geomorfol�ogico en el valle del río Oiartzun(Gipuzkoa). CKQ Estudios de Cuaternario 3, 19e34.

Dorado, J., Maeztu, J.J., Moreno, J., 2013. La catalogaci�on de cavidades en la CAPV.Karaitza 21, 22e35. www.euskalespeleo.com.

Era~na, C., Ugarte, F.M., 1992. Katabera, Sierra d'Aizkorri. Un karst sp�ecifique d'alti-tude en moyenne montagne basco-catabrique (Espagne). In: Salomon, J.N.,Maire, R. (Eds.), Karst et �evolutions climatiques. Universitaires de Bordeaux,pp. 309e316.

EVE, 1992. Mapa geol�ogico del País Vasco 1:25.000. Ente Vasco de Energía.Feuill�ee, P., Rat, P., 1971. Structures et paleog�eographies pyreneo-cantabriques. In:

Historia Structurales du Golfe de Gascogne, vol. 1, pp. 1e48.Flor, G., 1983. Las rasas asturianas: ensayos de correlaci�on y emplazamiento. Tra-

bajos de Geologia. Universidad de Oviedo 13, 65e81.Ford, D., Williams, P., 2007. Karst Hydrogeology and Geomorphology. Wiley

Chichester.Fuller, I.C., Macklin, M.G., Lewin, J., Passmore, D.G., Wintle, A.G., 1998. River

response to high-frequency climate oscillations in southern Europe over thepast 200 ky. Geology 26, 275e278.

Gabrov�sek, F., Dreybrodt, W., 2001. A model of the early evolution of karst aquifersin limestone in the dimensions of length and depth. Journal of Hydrology 240,206e224.

Gal�an, C., Del Cura, J.I., Lazkano, J., Nieto, M., 2004. Espeleología física del bloquek�arstico de Zelaieta (macizo de Izarraitz, País Vasco). Sociedad de CienciasAranzadi, pp. 1e29.

García-Mond�ejar, J., Hines, F.M., Pujalte, V., Reading, H.G., 1985. Sedimentation andtectonics in the western Basque-Cantabrian area (Northern Spain) duringcretaceous and Tertiary times. In: Mil�a, M.D., Rosell, J. (Eds.), 6th EuropeanRegional Meeting Excursion Guidebook, pp. 309e392.

García-Ruíz, J.M., Moreno, A., Gnz�alez-Samp�eriz, P., Valero-Garc�es, B., Martí-Bono, C., 2010. La cronología del último ciclo glaciar en las monta~nas del sur deEuropa. Una revisi�on. Cuaternario y Geomorfología 24, 35e46.

Granger, D., Fabel, D., Palmer, A., 2001. Pliocene-Pleistocene incision of the GreenRiver, Kentucky, determined from radioactive decay of cosmogenic 26Al and10Be in Mammoth Cave sediments. Geological Society of America Bulletin 113(7), 825e836.

Hallstadius, L., 1984. A method for the electrodeposition of actinides. Nuclear In-struments and Methods in Physics Research 223, 266e267.

H€auselmann, P., 2002. Cave Genesis and its Relationship to Surface Processes: In-vestigations in the Siebenhengste Region (BE, Switzerland). Unpublished PhDthesis. University of Fribourg, p. 168.

Hughes, P.D., 2010. Geomorphology and Quaternary stratigraphy: the roles ofmorpho-, and allostratigraphy. Geomorphology 123, 189e199.

Iriarte, E., S�anchez, M.A., Foyo, A., Tomillo, C., 2010. Geological risk assessment forcultural heritage conservation in karst caves. Journal of Cultural Heritage 11,250e258.

Ivanovich, M., Harmon, R.S., 1992. Uranium-series Disequilibrium: Applications toEarth, Marine, and Environmental Sciences. Clarendon Press, Oxford, p. 910.

Jim�enez-S�anchez, M., Bischoff, J., Stoll, H., Aranburu, A., 2006. A geochronologicalapproach for cave evolution in the Cantabrian Coast (Pindal Cave, NW Spain).Zeitschrift fur Geomorphologie 147, 129e141.

Jim�enez-S�anchez, M., Stoll, H., Giralt, S., Aranburu, A., Moreno, A., Domínguez-Cuesta, M.J., M�endez-Vicente, A., Ballesteros, D., Pirla, G., Valero-Garc�es, B.,Cheng, H., Edwards, L., 2011a. Marine terrace uplifting rates in the Cantabrianshore: contribution of U-Th speleothem dating. In: Deformation Mechanisms,Rheology and Tectonics DRT Meeting, Oviedo (Spain).

Jim�enez-S�anchez, M., Stoll, H., Giralt, S., Moreno, A., Domínguez-Cuesta, M.J.,Aramburu, A., M�endez-Vicente, A., Ballesteros, D., Valero, B., 2011b. Cave geo-morphology: a multidisciplinary approach in El Pindal Cave (NW Spain). In: 6thKarst Record Conference (KR6) on Climate Change. Birmingham, UK.

Kaufmann, G., 2009. Modelling karst geomorphology on different time scales.Geomorphology 106, 62e77.

Leorri, E., Cearreta, A., Milne, G., 2012. Field observations and modelling of Holocenesea-level changes in the southern Bay of Biscay: implication for understandingcurrent rates of relative sea-level change and vertical land motion along theAtlantic coast of SW Europe. Quaternary Science Reviews 42, 59e73.

Lewis, C.J., McDonald, E.V., Sancho, C., Pe~na, J.L., Rhodes, E.J., 2009. Climatic impli-cations of correlated Upper Pleistocene glacial and fluvial deposits on the Cincaand G�allego Rivers (NE Spain) based on OSL dating and soil stratigraphy. Globaland Planetary Change 67, 141e152.

Lisiecki, L.E., Raymo, M.E., 2005. A Pliocene-Pleistocene stack of 57 globallydistributed benthic d18O records. Paleoceanography 20, PA 1003.

Maiztegi, G., Rigault, P., Ugarte, L.F., Ugarte, F.M., 1974. Trabajos sobre el karst del SWde Gipuzkoa. Biblioteca municipal O~nati 5, p. 101.

Mary, G., 1983. Evoluci�on del margen costero de la Cordillera Cant�abrica en Asturiasdesde el Mioceno. Trabajos de Geología 13, 3e35.

Moreno, A., Stoll, H., Jim�enez-S�anchez, M., Cacho, I., Valoer-Garc�es, B., Ito, E.,Edwards, R., 2010. A speleothem record of glacial (25-22.6 kyr BP) rapid climaticchanges from northern Iberian Peninsula. Global and Planetary Change 71,218e231.

Moreno, A., Gonz�alez-Samp�eriz, P., Morell�on, M., Valero-Garc�es, B.L., Fletcher, W.J.,2012. Northern Iberian abrupt climate change dynamics during the last glacialcycle: a view from lacustrine sediments. Quaternary Science Reviews36,139e153.

Moreno, A., Belmonte, A., Bartolom�e, M., Sancho, C., Oliva, B., Stoll, H., Edwards, L.R.,Cheng, H., Hellstrom, J., 2013. Formaci�on de espeleotemas en el noreste

A. Aranburu et al. / Quaternary International 364 (2015) 217e230230

peninsular y su relaci�on con las condiciones clim�aticas durante los últimosciclos glaciares. Cuadernos de Investigaci�on Geogr�afica 25e47.

Palmer, A.N., 1987. Cave Levels and Their Interpretations, vol. 49. National Speleo-logical Society, pp. 50e66.

Palmer, A.N., 1991. Origin and morphology of limestones caves. Geological Society ofAmerica Bulletin 103, 1e21.

Pedoja, K., Husson, L., Regard, V., Cobbold, P.R., Ostanciaux, E., Johnson, M.E.,Kershaw, S., Saillard, M., Martinod, J., Furgerot, L., Weill, P., Delcaillau, B., 2011.Relative sea-level fall since the last interglacial stage: are coasts upliftingworldwide? Earth-Science Reviews 108, 1e15.

Pedoja, K., Husson, L., Johnson, M.E., Melnick, D., Witt, C., Pochat, S., Nexer, M.,Delcaillau, B., Pinegina, T., Poprawski, Y., Authemayou, C., Elliot, M., Regard, V.,Garestier, F., 2014. Coastal staircase sequences reflecting sea-level oscillationsand tectonic uplift during the Quaternary and Neogene. Earth-Science Reviews132, 13e38.

Pe~na, J.L., Sancho, C., Lewis, C., McDonald, E., Rhodes, E., 2004. Datos cronol�ogicos delas morrenas terminales del glaciar del G�allego y su relaci�on con las terrazasfluvioglaciares (Pirineo de Huesca). In: Pe~na, J.L., Longares, L.A., S�anchezFabre, M. (Eds.), El Medio Físico de Arag�on, Aspectos generales y tem�aticos.Universidad de Zaragoza e Instituci�on Fernando el Cat�olico, Zaragoza,pp. 71e84.

Pe~nalver, X., Mujika, J.A., 2005. Praile Aitz I (Deba, Gipuzkoa): evidenciasarqueol�ogicas y organizaci�on espacial en un suelo magdaleniense. In: Bicho, N.,Paleolítico, O. (Eds.), Actas do IV Congresso de Arqueología Peninsular, Pro-motoria Monogr�afica 2, pp. 143e156.

Rabineau, M., Bern�e, S., Olivet, J.L., Aslanian, D., Guillocheau, F., Joseph, P., 2006.Earth and Planetary Science Letters 252, 119e137.

Raymo, M.E., Mitrovica, J.X., O0Leary, M.J., DeConto, R.M., Hearty, P.J., 2011. De-partures from eustasy in Pliocene sea-level records. Nature geoscience 4,328e332.

Richards, D.A., Dorale, J.A., 2003. Uranium-series chronology and environmentalapplications of speleothems. Reviews in Mineralogy and Geochemistry 52,407e460.

Rodríguez-Rodríguez, L., Jim�enez-S�anchez, M., Domínguez-Cuesta, M.J.,Aranburu, A., 2015. Research history on glacial geomorphology and geochro-nology of the Cantabrian Mountains, north Iberia (43e42oN/7e2oW). Quater-nary International 364, 6e21. http://dx.doi.org/10.1016/j.quaint.2014.06.007.

Rosenbauer, R.J., 1991. UDATE1: a computer program for the calculation of Uranium-series isotopic ages. Computers & Geosciences 17, 45e75.

Sasowsky, I.D., 2007. Clastic sediments in caves e imperfect recorders of processesin karst. In: Kranjc, A., Gabrov�sek, F., Culver, D.C., Sasowsky, I.D. (Eds.), Time inKarst: Special Publication 12. Karst Waters Institute, Leesburg, Va, pp. 143e149.

Shackleton, N.J., S�anchez-Go~ni, M.F., Pailler, D., Lancelot, Y., 2003. Marine IsotopeSubstage5eandtheEemian Interglacial.Global andPlanetaryChange36,151e155.

Strasser, M., Strasser, A., Pelz, K., Seyfried, H., 2009. A mid Miocene to early Pleis-tocene multi-level cave as a gauge for tectonic uplift of the Swabian Alb(Southwest Germany). Geomorphology 106, 130e141.

Talvitie, N.A., 1972. Electrodeposition of actinides for alpha spectrometric deter-mination. Analytical Chemistry 44, 280e283.

Ugarte, F.M., 1985. La siera d'Aralar-mendi. Un exemple de karst de moyenne-montagne vasco cantabrique, vol. 903. Travauz U.A. du C.N.R.S. (Inst. Geo-graphie), pp. 55e76.

Ugarte, F.M., 1989. Geomorfología de las unidades k�arsticas situadas en los montesvascos. In: Dur�an, J.J., L�opez, J. (Eds.), El karst en Espa~na. Sociedad Espa~nola deGeomorfologia, pp. 151e163.

Ugarte, F.M., 1994. Montes Vasco-Cant�abricos. In: Gutierrez-Elorza, M. (Ed.), Geo-morfologia de Espa~na. Rueda, Madrid, pp. 227e250.

Vanara, N., 2000. Le karst des Arbailles, vol. 8. Karstologia M�emoire, p. 320.Waltham, T., 2009. Fengcong, fenglin, cone karst and tower karst. Cave and Karst

Science 35 (3), 77e88.